Combined ramjet control rule optimization method based on meta-heuristic calculation

Abstract

The invention provides a combined ramjet control rule optimization method based on meta-heuristic calculation, and belongs to the technical field of aero-engine control, and the method comprises the steps: building a combined ramjet model, and determining an optimization variable; according to the control mode of each combined ramjet, constraint conditions and an objective function based on optimization variables are determined, the control modes comprise a steady-state control mode and an acceleration control mode, and the steady-state control mode comprises a minimum fuel consumption control mode, a maximum thrust control mode and a minimum turbine front temperature control mode; and on the basis of the target function and the constraint condition, a meta-heuristic calculation strategy is adopted to optimize the control rule of each control mode, an optimal control variable is obtained, and the meta-heuristic calculation strategy is set to be a reptile search algorithm fused with a multi-dimensional learning prey strategy. According to the treatment scheme, the convergence speed of control rule optimization of the combined ramjet is increased, and the optimization effect and the engine performance are improved.

Description

Technical Field

[0001] The present application relates to the technical field of aeroengine control, and particularly relates to an optimization method for a combined ramjet engine control law based on metaheuristic calculation. Background Art

[0002] The combined ramjet engine, as the core power device of a new generation of hypersonic aircraft, has strategic significance in the fields of transonic flight and space-to-air round-trip. To meet the complex working condition requirements of a wide speed range and multiple modes (subsonic combustion / supersonic combustion ramjet, turbine-ramjet combination), it is crucial to optimize the comprehensive performance of the engine under mode conversion, thermodynamic cycle matching, and extreme aerodynamic heat environment. The engine control law can achieve multi-objective optimization such as thrust continuity, optimal specific impulse, and thermal protection by coordinating parameters such as the inlet adjustment mechanism, fuel distribution strategy, and combustion chamber mode switching. However, despite the breakthrough progress in the field of combined power control, the performance degradation caused by factors such as high-temperature ablation, shock wave interference, and inlet starting failure during long-term operation still poses a severe challenge. The traditional control law is difficult to adapt to the dynamically changing flow field characteristics, resulting in thrust oscillation and unstable mode conversion.

[0003] In recent years, metaheuristic optimization algorithms have gradually received extensive attention due to their unique solution mechanisms. For example, the particle swarm optimization algorithm, the grasshopper optimization algorithm, and the whale optimization algorithm have been applied to the optimization problems of complex systems. The reptile algorithm (RSA) can also be changed to other animal algorithms, and its formula is modified. For example, a metaheuristic optimization algorithm driven by the hunting behavior of crocodiles has a clear structure, simple implementation, and good stability, so it has received attention. However, similar to other metaheuristic optimization algorithms, RSA also faces problems such as a large demand for key length, high storage and transmission costs, and time-consuming key generation processes. Therefore, it is of great significance for the control law optimization technology to study an optimization strategy that can solve the performance degradation problem during the long-term operation of aeroengines. Summary of the Invention

[0004] In view of this, the embodiments of the present application provide an optimization method for a combined ramjet engine control law based on metaheuristic calculation, which at least partially solves the problems that the engines in long-term service in the prior art cannot adaptively adjust after component degradation and the optimization accuracy of traditional optimization algorithms is low.

[0005] The embodiments of the present application provide an optimization method for a combined ramjet engine control law based on metaheuristic calculation, and the method includes: Establish a combined ramjet engine model, and determine optimization variables based on the combined ramjet engine model; For each control mode of the combined ramjet engine, constraint conditions and an objective function based on optimization variables are determined. The control modes include a steady-state control mode and an acceleration control mode. The steady-state control mode includes a minimum fuel consumption control mode, a maximum thrust control mode, and a minimum pre-turbine temperature control mode; Based on the objective function and the constraint conditions, a meta-heuristic calculation strategy is used to optimize the control law of each control mode to obtain optimal control variables. The meta-heuristic calculation strategy is set as a reptile search algorithm fused with a multi-dimensional learning hunting strategy.

[0006] According to a specific implementation manner of an embodiment of the present application, the optimization variables are: , The constraint conditions in the process of optimizing the control law of each control mode are described as: , where u is the optimization variable, W f is the fuel flow rate of the turbine engine combustion chamber, A et is the throat area of the turbine engine nozzle, φ is the fuel equivalence ratio of the ramjet engine combustion chamber, A nt is the throat area of the ramjet engine nozzle, AI tur is the opening of the turbine flow passage, AI tam is the opening of the ramjet flow passage, A it is the throat area of the ramjet engine inlet, T4 is the total temperature at the combustion chamber outlet section, ζ sm is the surge margin of the turbine engine compressor, n is the compressor speed, ζ inlet is the inlet margin, π inlet is the inlet pressure ratio of the ramjet engine, P3 is the compressor outlet pressure, the subscript min is the minimum value, and the subscript max is the maximum value.

[0007] According to a specific implementation manner of an embodiment of the present application, the objective function and constraint conditions of the minimum fuel consumption control mode are expressed as: , where min sfc is the minimum value of the engine fuel consumption rate, f(u) is the objective function, s.t. is the constraint condition, g i’ (u) is the inequality constraint condition, h k’ (u) is the equality constraint condition, i’ is the i’th inequality constraint condition, I’ is the total number of inequality constraint conditions, k’ is the k’th equality constraint condition, and K’ is the total number of equality constraint conditions; The objective function and constraint conditions of the maximum thrust control mode are expressed as: , Among them, max F is the maximum engine thrust; The objective function and constraint conditions of the lowest turbine inlet temperature control mode are expressed as: .

[0008] According to a specific implementation manner of an embodiment of the present application, the method further includes: converting the constrained problem of the control law into an unconstrained problem by using a penalty function, and the expression of the penalty function is: , Among them, F(u,σ) is the objective function introducing the penalty function, and σ1 and σ2 are positive infinity.

[0009] According to a specific implementation manner of an embodiment of the present application, the reptile search algorithm integrating the multi-dimensional learning hunting strategy includes the following steps: Initialize the reptile search algorithm to generate a candidate solution set; Based on the candidate solution set, update the position in the reptile surrounding stage; Based on the hunting coordination and hunting cooperation strategies, update the position in the hunting stage; Update the reptile position based on the multi-dimensional learning hunting strategy to improve the quality of search individuals and increase the search ability, and obtain the current best position and the current best fitness value; Evaluate the fitness value of the individual at the current best position by using the greedy strategy, and retain the individuals valuable for the population position update to obtain the final global optimal position and the best fitness value.

[0010] According to a specific implementation manner of an embodiment of the present application, the expression of the set of candidate solutions is: , , Among them, x is the set of candidate solutions, and x i,j represents the j th th position of the i th th solution, N represents the number of candidate solutions, n represents the dimension size of the given problem, rand is a random number from 0 to 1, LB is the lower bound of the given problem, and UB is the upper bound of the given problem.

[0011] According to a specific implementation manner of an embodiment of the present application, the position update equation in the surrounding stage is: , The position update equation in the hunting stage is: , Among them, , , , , , Among them, x i,j (t+1) indicates the i-th th The jth solution th The location after the location update, Best j (t) is the jth optimal solution so far. th positions, t is the number of current iterations, T is the maximum number of iterations, η (i,t) (t) is the i-th th The jth solution th The hunting operator at each position, β is the first sensitive parameter, α is the second sensitive parameter, R (i,j) (t) is the value used to reduce the search area, ES(t) is the probability ratio, and ES(t) randomly takes a decreasing value between 2 and -2 in the number of iterations. ∈ represents infinitesimal, r1 is a random number between [1N], r2 is a random number between [1N], r3 is a random integer between -1 and 1, P (i,j) is the percentage difference between the optimal solution position and the current solution position, M(x i ) is i th The average position of the solution, UB (j) For the i th The upper boundary of the solution, LB (j) For the i th The lower bound of a solution is ∈ ' is an infinitesimal positive integer.

[0012] According to a specific implementation of the embodiment of the present application, the updating of the position of the reptile based on the multi-dimensional learning hunting strategy includes: constructing a radius matrix based on the original position of the reptile and the new position of the reptile; Construct a neighborhood matrix based on the radius matrix and the Euclidean distance between the current individual and the candidate individuals; New individuals are generated by learning from multiple neighborhood matrices, where the d-th dimension of each new individual is based on a randomly selected reptile position.

[0013] According to a specific implementation of the embodiment of the present application, the radius matrix expression is: , The neighborhood matrix expression is: , The expression of the generated new individual is: , where Radiusi(t) is the radius matrix, Neighbouri(t) is the neighborhood matrix, x i (t) is the current individual, x j (t) is the alternative individual, x new (t + 1) is the new position of individual i at the (t + 1)-th iteration, D is the spatial dimension of the optimization variable, N is the initial population, x i-DLH,j (t + 1) is the generated new individual, x i,d (t) is the updated position of individual i in the d-th dimension at the t-th iteration, x n,d (t) is the individual randomly selected by individual i in the d-th dimension at the t-th iteration, x r,d (t) is the reference position of individual i in the d-th dimension at the t-th iteration.

[0014] According to a specific implementation manner of the embodiment of the present application, the expression of the greedy strategy is: , where x i (t + 1) is the updated position of individual i at the (t + 1)-th iteration, x i-new (t + 1) is the position of individual i after the (t + 1)-th iteration without executing the multi-dimensional learning hunting strategy, x i-DLH (t + 1) is the position of individual i after the (t + 1)-th iteration after the multi-dimensional learning hunting strategy is updated, f(x i-new ) is the objective function corresponding to x i-new , f(x i-DLH ) is the objective function corresponding to x i-DLH .

[0015] Beneficial effects: In the method for optimizing the control law of a combined ramjet based on meta-heuristic calculation in the embodiment of the present application, this method improves the reptile search optimization algorithm, adopts a multi-dimensional learning hunting strategy for optimizing the steady-state and acceleration control laws of the combined ramjet, improves the optimization effect of the control law and enhances the performance of the engine. The simulation results show that the multi-dimensional learning hunting strategy combined with the reptile search algorithm has a faster convergence speed and better optimization effect than several classical intelligent optimization algorithms. It effectively reduces the fuel consumption rate and the minimum turbine inlet temperature of the combined ramjet, increases the maximum thrust, and shortens the acceleration time by 50%. Description of the Drawings

[0016] To more clearly illustrate the technical solutions of the embodiments of the present application, the accompanying drawings required for the embodiments will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can be obtained based on these drawings.

[0017] Figure 1 It is a flowchart of an optimization method for the combined ramjet control law based on meta-heuristic calculation according to an embodiment of the present invention. Detailed implementation manners

[0018] The embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0019] The following uses specific specific examples to illustrate the implementation manners of the present application. Those skilled in the art can easily understand other advantages and effects of the present application from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The present application can also be implemented or applied through other different specific implementation manners. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that, without conflict, the following embodiments and the features in the embodiments can be combined with each other. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts belong to the scope of protection of the present application.

[0020] It should be noted that the following describes various aspects of the embodiments within the scope of the appended claims. It should be obvious that the aspects described herein can be embodied in a wide variety of forms, and any specific structure and / or function described herein is only illustrative. Based on the present application, those skilled in the art should understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects described herein can be used to implement the device and / or practice the method. In addition, this device and / or this method can be implemented using other structures and / or functions in addition to one or more of the aspects described herein.

[0021] It should also be noted that the diagrams provided in the following embodiments only schematically illustrate the basic concept of the present application. The diagrams only show the components related to the present application and are not drawn according to the number, shape, and size of the components in actual implementation. The type, quantity, and ratio of each component in its actual implementation can be an arbitrary change, and the component layout type may also be more complex.

[0022] In addition, in the following description, specific details are provided to facilitate a thorough understanding of the examples. However, those skilled in the art will understand that the described aspects can be practiced without these specific details.

[0023] An embodiment of the present application provides an optimization method for the control law of a combined ramjet engine based on meta-heuristic calculation. The following is a detailed description with reference to Figure 1 this.

[0024] In one embodiment, an optimization method for the control law of a combined ramjet engine based on meta-heuristic calculation, the method includes the following steps: Establish a combined ramjet engine model, and determine the optimization variables based on the combined ramjet engine model; For each control mode of the combined ramjet engine, determine the constraint conditions and the objective function based on the optimization variables. The control modes include a steady-state control mode and an acceleration control mode. The steady-state control mode includes a minimum fuel consumption control mode, a maximum thrust control mode, and a minimum pre-turbine temperature control mode; Based on the objective function and the constraint conditions, use a meta-heuristic calculation strategy to optimize the control law of each control mode to obtain the optimal control variables. The meta-heuristic calculation strategy is set as a reptile search algorithm fused with a multi-dimensional learning hunting strategy.

[0025] In this embodiment, the meta-heuristic calculation strategy is set as a reptile search algorithm fused with a multi-dimensional learning hunting strategy (DLH-IRSA). The multi-dimensional learning hunting strategy (DLH, Dimensional Learning Harvesting) fuses an improved reptile search algorithm (RSA, Reptile Search Algorithm). The reptile search algorithm is a heuristic optimization algorithm based on the foraging behavior of reptiles in nature.

[0026] In this embodiment, the proposed control law does not directly optimize the engine control parameters, but optimizes the controlled parameters (such as thrust and specific impulse) and other adjustment parameters. Compared with directly optimizing the control parameters, adjusting the controlled parameters has the following advantages: First, since the engine is easily affected by external factors (such as flight altitude, gusts, etc.), if the control parameters are directly optimized, the control parameters need to be adjusted in real time according to the environment and flight state, and it is difficult for existing optimization algorithms to meet the requirements of both accuracy and real-time performance. Second, since the engine performance parameters cannot be directly measured, the usual control system cannot directly control the performance parameters. This embodiment proposes an improvement to the reptile search optimization algorithm, adopting a multi-dimensional learning hunting strategy for optimizing the steady-state and acceleration control laws of the combined ramjet engine, improving the optimization effect of the control law and enhancing the performance of the engine. The simulation results show that the multi-dimensional learning hunting strategy combined with the reptile search algorithm has a faster convergence speed and better optimization effect than several classical intelligent optimization algorithms. It effectively reduces the fuel consumption rate and the lowest turbine inlet temperature of the combined ramjet engine, increases the maximum thrust, and shortens the acceleration time by 50%.

[0027] Furthermore, for the combined ramjet engine model studied in this embodiment, the optimization variables are as follows: (1), The constraint conditions in the process of optimizing the control law for each control mode are described as follows: (2), where u is the optimization variable, W f is the fuel flow rate in the turbine engine combustion chamber, A et is the throat area of the turbine engine nozzle, φ is the fuel equivalence ratio in the ramjet engine combustion chamber, A nt is the throat area of the ramjet engine nozzle, AI tur is the opening of the turbine flow passage, AI tam is the opening of the ramjet flow passage, A it is the throat area of the ramjet engine inlet, T4 is the total temperature at the combustion chamber outlet section, ζ sm is the surge margin of the turbine engine compressor, n is the compressor speed, ζ inlet is the inlet margin, π inlet is the inlet pressure ratio of the ramjet engine, P3 is the compressor outlet pressure, the subscript min represents the minimum value, and the subscript max represents the maximum value.

[0028] For the above expression of the constraint conditions, it can be expressed as: (3), where g i’ (u) is the inequality constraint condition, and i’ represents the i’th inequality constraint condition.

[0029] In one embodiment, in the minimum fuel consumption control mode, the equality constraint emphasizes achieving the minimum fuel consumption rate while maintaining constant thrust. It is necessary to comprehensively adjust various parameters to improve the working efficiency of the fan, compressor and inlet, increase the engine flow, and thus achieve thrust improvement. By solving the nonlinear constraint problem, find a suitable combination of control variables so that the engine thrust reaches the expected value and runs at the minimum fuel consumption rate point. The objective function and constraint conditions of the minimum fuel consumption control mode are expressed as: (4), Among them, min sfc is the minimum value of the engine fuel consumption rate, f(u) is the objective function, st is the constraint condition, g i’ (u) is the inequality constraint, h k’ (u) is the equality constraint, i' is the i'th inequality constraint, I' is the total number of inequality constraints, k' is the k'th equality constraint, and K' is the total number of equality constraints; The maximum thrust control mode is mainly used in high thrust demand stages such as aircraft takeoff and accelerated climb. In this mode, the goal is to achieve maximum thrust while ensuring safe engine operation. At the turbine engine operating point, the maximum thrust is located at the intersection of the safety boundaries of the minimum compressor surge margin and the highest total temperature at the combustion chamber outlet; at the ramjet engine operating point, the maximum thrust is located at the intersection of the safety boundaries of the minimum inlet surge margin and the highest total temperature at the combustion chamber outlet; in the mode transition state, the maximum thrust is at the intersection of the safety boundaries of the inlet and compressor margin and the highest total temperature at the combustion chamber outlet. The objective function and constraints of the maximum thrust control mode are expressed as: (5), Among them, max F is the maximum value of engine thrust; The minimum turbine inlet temperature control mode is mainly used in high Mach number flight conditions of turbine engines. In this mode, the goal is to reduce the turbine inlet temperature as much as possible to extend the engine service life and reduce infrared radiation while ensuring that the thrust remains unchanged. This mode mainly reduces the engine turbine temperature by reducing the main fuel flow rate, and improves the engine compression ratio, efficiency and flow rate by comprehensively adjusting parameters such as the tail nozzle area, the fan guide vane angle and the compressor guide vane angle, thereby increasing the thrust and ultimately keeping the thrust basically constant. The objective function and constraints of the minimum turbine inlet temperature control mode are expressed as: (6).

[0030] In specific implementation, both the optimization of the minimum fuel consumption control mode and the optimization of the minimum turbine inlet temperature mode are multi-objective optimizations. The objective of the minimum fuel consumption control mode optimization is to achieve the expected thrust while minimizing the specific fuel consumption; the objective of the minimum turbine inlet temperature mode optimization is to achieve the expected thrust while minimizing the turbine inlet temperature. By gradually iterative optimization, either objective 1 or objective 2 is made optimal. By selecting an appropriate weight W = [W1, W2], where W1 is the weight of objective 1 (minimum specific fuel consumption, minimum turbine inlet temperature) and W2 is the weight of objective 2 (constant thrust), the comprehensive performance of objective 1 and objective 2 is made optimal.

[0031] Furthermore, by combining formulas (4) to (6), the optimization of the combined ramjet control law can be expressed by the following non-linear programming problem: (7), To solve the constrained optimization problem, a penalty function can be introduced. (8), where F(u, σ) is the objective function with the penalty function introduced, and σ1 and σ2 are infinitely large positive numbers, thus converting the original constrained problem into an unconstrained problem. In the intelligent optimization process of the combined ramjet control law, an intelligent optimization algorithm is needed to find the best combination of optimization parameters to make the engine performance optimal.

[0032] In one embodiment, a multi-dimensional learning hunting strategy fusion reptile search optimization algorithm is proposed. By this method, DLH-IRSA overcomes the defects of the original RSA, further improves the optimization speed and solution accuracy, and provides an effective strategy for solving complex problems. The reptile search algorithm fusion multi-dimensional learning hunting strategy includes the following steps: Initialize the reptile search algorithm to generate a candidate solution set; Based on the candidate solution set, update the position in the reptile surrounding stage; Based on the hunting coordination and hunting cooperation strategies, update the position in the hunting stage; Update the reptile position based on the multi-dimensional learning hunting strategy to improve the quality of search individuals and increase the search ability, and obtain the current best position and the current best fitness value; Adopt a greedy strategy to evaluate the fitness value of the individuals at the current best position, and retain the individuals valuable for the population position update to obtain the final global optimal position and the best fitness value.

[0033] Furthermore, the expression of the set of candidate solutions is: , , Among them, x is the set of candidate solutions, and x i,j represents the j th -th position of the i th -th solution. N represents the number of candidate solutions, n represents the dimension size of the given problem, rand is a random number between 0 and 1, LB is the lower bound of the given problem, and UB is the upper bound of the given problem.

[0034] Furthermore, the position update equation in the encircling stage is: , and the position update equation in the hunting stage is: , where , , , , , where x i,j (t + 1) represents the updated position of the j th -th position of the i th -th solution, Best j (t) is the j th -th position in the optimal solution so far, t is the current iteration number, T is the maximum iteration number, η (i,t) (t) is the hunting operator for the j th -th position in the i th -th solution, β is the first sensitivity parameter, α is the second sensitivity parameter, R (i,j) (t) is the value used to reduce the search area, ES(t) is the probability ratio, and ES(t) randomly takes a decreasing value between 2 and -2 during the iteration number, ∈ represents an infinitesimal, r1 is a random number between [1, N], r2 is a random number between [1, N], r3 is a random integer between -1 and 1, P (i,j) is the percentage difference between the position of the optimal solution and the position of the current solution, M(x i ) is the average position of the i th -th solution, UB (j) is the upper bound of the i th -th solution, LB (j) is the lower bound of the i th -th solution, ∈ ’ is an infinitesimal positive integer.

[0035] Furthermore, updating the positions of reptiles based on the multi-dimensional learning hunting strategy includes: Construct a radius matrix based on the original position and the new position of the reptile; Construct a neighborhood matrix based on the radius matrix and the Euclidean distance between the current individual and the alternative individual; Generate new individuals by learning from multiple neighborhood matrices, where the d-th dimension of each new individual is based on randomly selected reptile positions.

[0036] Furthermore, the expression of the radius matrix is: , The expression of the neighborhood matrix is: , The expression of the generated new individual is: , where Radiusi(t) is the radius matrix, Neighbouri(t) is the neighborhood matrix, x i (t) is the current individual, x j (t) is the alternative individual, x new (t + 1) is the new position of individual i at the (t + 1)-th iteration, D is the dimensionality of the optimization variables, N is the initial population, x i-DLH,j (t + 1) is the generated new individual, x i,d (t) is the updated position of individual i in the d-th dimension at the t-th iteration, x n,d (t) is the individual randomly selected by individual i in the d-th dimension at the t-th iteration, x r,d (t) is the reference position of individual i in the d-th dimension at the t-th iteration.

[0037] Furthermore, the expression of the greedy strategy is: , where x i (t + 1) is the updated position of individual i at the (t + 1)-th iteration, x i-new (t + 1) is the position of individual i after the (t + 1)-th iteration without executing the multi-dimensional learning hunting strategy, x i-DLH (t + 1) is the position of individual i after the (t + 1)-th iteration after the update of the multi-dimensional learning hunting strategy, f(x i-new ) is the objective function corresponding to x i-new ,f(x i-DLH ) is the objective function corresponding to x i-DLH .

[0038] The following uses a specific embodiment to describe in detail the reptile search algorithm integrated with the multi-dimensional learning hunting strategy, which specifically includes the following steps: Step 1: Initialization Phase In the RSA algorithm, the optimization process starts with a randomly generated set of candidate solutions (as shown in Equation (9)), and the optimal solution is considered the best solution in each iteration.

[0039] (9), where x is the set of candidate solutions randomly generated using Equation (10), as follows: (10).

[0040] Step 2: Encirclement Phase The search in this phase is conditional on two conditions. The movement strategy of walking on high places is conditional on and the movement strategy of walking on the abdomen is conditional on and . This means that this condition will be satisfied for almost half of the exploration iterations (walking on high places) and the other half of walking on the abdomen, which are two exploration search methods. Note that to check the random scaling factor for element generation of more diverse solutions and explore the diverse region, the most direct rule is adopted in this embodiment, which can mimic the encircling behavior of crocodiles. The position update equation for the exploration phase is proposed, as shown in Equation (11), (11), where, Best j (t) is the j th th position in the optimal solution so far, t is the number of the current iteration, T is the maximum number of iterations, η (i,t) (t) is the hunting operator at the j th th position in the i th th solution (calculated by Equation (12)), β is the first sensitive parameter that controls the detection accuracy of the encirclement phase during the iteration process (i.e., walking on high places), and is fixed at 0.1. R (i,j) (t) is the value used to reduce the search area (calculated by Equation (13)), and the evolutionary sense ES(t) is the probability ratio. ES(t) randomly takes a decreasing value between 2 and -2 during the iteration (calculated by Equation (14)).

[0041] (12), (13), (14), In Equation (14), 2 is an associated value, r2 is a random number between [1 N], r3 is a random integer between -1 and 1, and P (i,j) is the percentage difference between the position of the optimal solution and the position of the current solution, calculated by Equation (15); (15), (16), M(x i ) is the average position of the i th -th solution, calculated by formula (16), UB (j) is the upper bound of the i th -th solution, LB (j) is the lower bound of the i th -th solution, ∈ ' is an infinitesimal positive integer, and α is a sensitivity parameter, which also controls the exploration accuracy (the difference between candidate solutions) of hunting cooperation during the iteration process and is fixed at 0.1 in the embodiment.

[0042] Step 3: Hunting stage The mechanism of this algorithm is to search for the optimal solution through the search space and method, using two main search strategies (i.e., hunting coordination and hunting cooperation); the model is shown in formula (17). The search condition for this stage is that the condition for the hunting coordination strategy is and , otherwise, execute the hunting cooperation strategy when and . Note that the random coefficient is considered to generate denser solutions and utilize promising regions (locally). The most straightforward rule is adopted in this embodiment, which can imitate the hunting behavior of crocodiles. The following position update equation is proposed for the development stage in this embodiment: (17), In this regard, when , the cyclic stage (exploration) occurs, otherwise, when , the hunting stage (exploitation) occurs.

[0043] The exploitation search mechanisms (hunting coordination and cooperation) attempt to avoid getting trapped in local optima. These processes help the exploration search to determine the optimal solution and maintain the diversity of candidate solutions. Two parameters (i.e., β and α) are designed in this application to generate a random value at each iteration, and continue to explore not only during the first iteration but also during the last iteration. This part of the search is beneficial in the case of local optimum stagnation, especially in the final iteration.

[0044] Step 4: Update the position using the DLH strategy Traditional RSA has the dilemma of low computational efficiency. Therefore, DLH is introduced to improve the quality of search individuals and increase the search ability.

[0045] DLH generates the following alternative individuals: (18), Among them, the radius matrix Radius is generated by subtracting the distance between the original and new positions, and the neighborhood matrix is constructed based on the Euclidean distance between the current individual x i (t) and the alternative individual x j (t).

[0046] (19), By learning from many neighborhood matrices, DLH generates a new individual x i-DLH,j (t + 1), where the d-th dimension of each new individual is updated based on the d-th dimension of a randomly selected individual x n,d (t).

[0047] (20), Recalculate the individual fitness value, and retain the current best position and the current best fitness value.

[0048] Finally, a greedy strategy is adopted to interfere with the current optimal position, evaluate the fitness value of the current optimal individual, and retain the individuals that are more valuable for updating the population position. The mathematical model is as follows: (21).

[0049] When the current iteration number does not satisfy t < t max , finally return the best fitness value and the global optimal position.

[0050] The present invention proposes a method for optimizing the control law of a combined ramjet engine based on meta-heuristic calculation, introduces a reptile search algorithm, and combines a multi-dimensional learning hunting strategy to improve the global convergence of the algorithm, aiming to achieve the lowest fuel consumption rate, thereby improving fuel efficiency; achieving the maximum thrust when the power demand is the largest to meet specific flight requirements; and maintaining the turbine inlet temperature at the lowest level to ensure the service life of the turbine; ensuring that the engine system parameters do not exceed the limit values during the transient process and achieving the shortest acceleration time. The specific contributions of the present invention are as follows: (1) Propose a multi-dimensional learning hunting strategy fusion reptile search algorithm (DLH-IRSA) to improve the optimization speed and solution accuracy of the algorithm. And by introducing a multi-dimensional learning hunting strategy, the individual quality is improved, the search process is enhanced, and the exploration and exploitation phases are balanced.

[0051] (2) Use the DLH-IRSA algorithm and a variety of classic intelligent optimization algorithms to conduct numerical simulation verification on the intelligent optimization of the combined ramjet engine control law to confirm the feasibility of the proposed method. Select the optimization parameters as the optimization variables, and optimize for the steady-state control mode and the transient control mode respectively.

[0052] The following describes the optimization method of the combined ramjet engine control law based on meta-heuristic calculation of the present application with a specific embodiment.

[0053] The DLH-IRSA algorithm and a variety of classical intelligent optimization algorithms are used to conduct numerical simulation verification on the intelligent optimization of the control law to prove the feasibility of the proposed method. The optimization parameters are selected. Optimization is carried out for the steady-state control mode and the acceleration control mode respectively.

[0054] (1) Minimum fuel consumption control mode At low Mach numbers, the combined ramjet engine adopts the turbine operating mode. Taking the subsonic cruise point as an example, the minimum fuel consumption control method of the combined ramjet engine is verified and verified by the combined ramjet engine. The optimization objective function is set as , where fitness is the fitness, F r is the reference thrust of the engine, and F n is the actual thrust of the engine.

[0055] (2) Maximum thrust control mode In the ramjet engine mode, taking the high Mach constant acceleration flight as an example, the maximum thrust control method of the combined ramjet engine is simulated. The optimization objective function is set as fitness = 10000 / F n .

[0056] (3) Minimum turbine inlet temperature control mode The minimum turbine inlet temperature control of the combined ramjet engine is usually applied to high-altitude high Mach flight. The goal of the minimum turbine inlet temperature control is to ensure that the gas temperature in front of the turbine does not exceed the maximum allowable temperature of the turbine material, thereby protecting the turbine. Taking supersonic flight as an example, the intelligent optimization algorithm proposed in the present invention and various classical intelligent optimization algorithms are used to optimize the minimum turbine inlet temperature control mode. The optimization objective function is set as fitness = (F r - F n ) 2 + T4.

[0057] In the embodiment provided by the present invention, the method improves the reptile search optimization algorithm and adopts a multi-dimensional learning hunting strategy for the steady-state and acceleration control law optimization of the combined ramjet engine, improving the optimization effect of the control law and enhancing the performance of the engine. The simulation results show that the multi-dimensional learning hunting strategy combined with the reptile search algorithm has a faster convergence speed and better optimization effect than several classical intelligent optimization algorithms. It effectively reduces the fuel consumption rate and the minimum turbine inlet temperature of the combined ramjet engine, increases the maximum thrust, and shortens the acceleration time by 50%.

[0058] As described above, it is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed in the present application should be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A method for optimizing the control law of a combined ramjet engine based on meta-heuristic calculation, characterized in that The method includes: Establishing a combined ramjet engine model and determining optimization variables based on the combined ramjet engine model; For each control mode of the combined ramjet engine, determining constraint conditions and an objective function based on the optimization variables, where the control modes include a steady-state control mode and an acceleration control mode, and the steady-state control mode includes a minimum fuel consumption control mode, a maximum thrust control mode, and a minimum pre-turbine temperature control mode; Based on the objective function and the constraint conditions, using a metaheuristic calculation strategy to optimize the control law of each control mode to obtain optimal control variables, and the metaheuristic calculation strategy is set as a reptile search algorithm fused with a multi-dimensional learning hunting strategy.

2. The optimization method of the combined ramjet control law based on meta-heuristic calculation according to claim 1, characterized in that The optimization variables are: ， The constraint conditions in the process of optimizing the control law of each control mode are described as: ， where, u is the optimization variable, W f is the fuel flow rate of the turbine engine combustion chamber, A et is the throat area of the turbine engine nozzle, φ is the fuel equivalence ratio of the ramjet combustion chamber, A nt is the throat area of the ramjet nozzle, AI tur is the opening of the turbine flow passage, AI tam is the opening of the ramjet flow passage, A it is the throat area of the ramjet inlet, T4 is the total temperature at the combustion chamber outlet section, ζ sm is the surge margin of the turbine engine compressor, n is the compressor speed, ζ inlet is the inlet margin, π inlet is the inlet pressure ratio of the ramjet, P3 is the compressor outlet pressure, the subscript min is the minimum value, and the subscript max is the maximum value.

3. The method for optimizing the control law of a combined ramjet engine based on metaheuristic calculation according to claim 2, characterized in that The objective function and constraint conditions of the minimum fuel consumption control mode are expressed as: ， Among them, min sfc is the minimum value of the engine fuel consumption rate, f(u) is the objective function, s.t. is the constraint condition, and g i’ (u) is the inequality constraint condition, and h k’ (u) is the equality constraint condition, i’ is the i’-th inequality constraint condition, I’ is the total number of inequality constraint conditions, k’ is the k’-th equality constraint condition, and K’ is the total number of equality constraint conditions; The objective function and constraint conditions of the maximum thrust control mode are expressed as: ， where max F is the maximum engine thrust; The objective function and constraint conditions of the minimum pre-turbine temperature control mode are expressed as: 。 4. The method for optimizing the combined ramjet control law based on meta-heuristic calculation according to claim 3, wherein The method further includes: using a penalty function to convert the constrained problem of the control law into an unconstrained problem, and the expression of the penalty function is: ， where F(u,σ) is the objective function with the penalty function introduced, and σ1 and σ2 are positive infinity numbers.

5. The optimization method for the combined ramjet control law based on metaheuristic calculation according to claim 1, wherein The reptile search algorithm fused with the multi-dimensional learning hunting strategy includes the following steps: Initializing the reptile search algorithm to generate a candidate solution set; Based on the candidate solution set, performing position update in the reptile surrounding stage; Based on the hunting coordination and hunting cooperation strategies, performing position update in the hunting stage; Updating the reptile position based on the multi-dimensional learning hunting strategy to improve the quality of search individuals and increase the search ability, and obtaining the current best position and the current best fitness value; Using a greedy strategy to evaluate the fitness value of the individuals at the current best position and retaining the individuals valuable for population position update to obtain the final global optimal position and the best fitness value.

6. The optimization method for the combined ramjet control law based on meta-heuristic calculation according to claim 5, characterized in that The expression of the set of candidate solutions is: ， ， Among them, \(x\) is the set of candidate solutions, and \(x_{ij}\) i,j represents the \(j\)th th position of the \(i\)th th solution, \(N\) represents the number of candidate solutions, \(n\) represents the dimensionality size of the given problem, \(rand\) is a random number from 0 to 1, \(LB\) is the lower bound of the given problem, and \(UB\) is the upper bound of the given problem.

7. The optimization method for the combined ramjet control law based on meta-heuristic calculation according to claim 6, wherein The position update equation in the surrounding stage is: ， The position update equation in the hunting stage is: ， where ， ， ， ， ， Among them, x i,j (t+1) indicates the i-th th The jth solution th The location after the location update, Best j (t) is the jth optimal solution so far. th positions, t is the number of current iterations, T is the maximum number of iterations, η (i,t) (t) is the i-th th The jth solution th The hunting operator at each position, β is the first sensitive parameter, α is the second sensitive parameter, R (i,j) (t) is the value used to reduce the search area, ES(t) is the probability ratio, and ES(t) randomly takes a decreasing value between 2 and -2 in the number of iterations. ∈ represents infinitesimal, r1 is a random number between [1N], r2 is a random number between [1N], r3 is a random integer between -1 and 1, P (i,j) is the percentage difference between the optimal solution position and the current solution position, M(x i ) is i th The average position of the solution, UB (j) For the i th The upper boundary of the solution, LB (j) For the i th The lower bound of a solution is ∈ ' is an infinitesimal positive integer.

8. The optimization method for the combined ramjet control law based on meta-heuristic calculation according to claim 7, wherein The updating of the reptile position based on the multi-dimensional learning hunting strategy includes: Constructing a radius matrix based on the original position of the reptile and the new position of the reptile; Constructing a neighborhood matrix based on the radius matrix and the Euclidean distance between the current individual and the alternative individual; Generating new individuals by learning from multiple neighborhood matrices, where the d-th dimension of each new individual is based on randomly selected reptile positions.

9. The optimization method for the combined ramjet control law based on meta-heuristic calculation according to claim 8, characterized in that The expression of the radius matrix is: ， The expression of the neighborhood matrix is: ， The expression of the generated new individual is: ， Among them, Radiusi(t) is the radius matrix, Neighbouri(t) is the neighborhood matrix, x i (t) is the current individual, x j (t) is the alternative individual, x new (t + 1) is the new position of individual i at the (t + 1)-th iteration, D is the spatial dimension of the optimization variable, N is the initial population, x i-DLH,j (t + 1) is the newly generated individual, x i,d (t) is the updated position of individual i at the d-th dimension at the t-th iteration, x n,d (t) is the individual randomly selected by individual i at the d-th dimension at the t-th iteration, x r,d (t) is the reference position of individual i at the d-th dimension at the t-th iteration.

10. The optimization method for the combined ramjet control law based on metaheuristic calculation according to claim 9, wherein The expression of the greedy strategy is: ， Among them, x i (t + 1) is the updated position of individual i at the (t + 1)-th iteration, x i-new (t + 1) is the position of individual i who has not executed the multi-dimensional learning hunting strategy after the (t + 1)-th iteration, x i-DLH (t + 1) is the position of individual i after the multi-dimensional learning hunting strategy is updated at the (t + 1)-th iteration, f(x i-new ) is the objective function corresponding to x i-new , f(x i-DLH ) is the objective function corresponding to x i-DLH .

Images